Spiro[isobenzofuran-1(3H),4’-piperidines]
Journal of Medicinal Chemistry, 1978, Vol. 21, No. 11 1149
deviation of 0.04, were made at two different concentrations and volumes of phases. Following shaking for 30 min, the layers were separated, the aqueous portion was centrifuged at 99OOg and then filtered, and absorbances were measured at 301,261, and 290 nm for 4,5, and 7, respectively. In addition, the log P value of 5 was determined to be 4 . 7 6 using an octanol-water system, with the concentration of 5 determined as the inorganic phosphate by treating the sample with a sulfuric acid-perchloric acid solution16 and quantitating phosphate concentrations using the method of Berenblum and Chain.” Syntheses. The hydrazones (starting hydrazides’ references cited) 2,’* 4-7, 10, and ll,l99,4 and 3 and 1220(Table 11) were prepared by refluxing the appropriate hydrazide with pyridine-2-carboxaldehyde or pyridine-2-carboxaldehyde l-oxide2’ in absolute EtOH for 1 h in the presence of HOAc (1 mL). Where no recrystallization solvent is indicated, the products were isolated from the reaction mixture, washed with water, and dried in vacuo. No acetic acid was used in the case of 2, 4, 7, or 9. Diphenyl phosphorochloridothionate, required for the preparation of the hydrazide leading to 3 and 12, was synthesized according to the procedure of Autenrieth and Hildebrand.*’ The MnOp required for the synthesis of pyridine-2-carboxaldehyde 1-oxide is best prepared by the method of Sondheimer et al.23 Cupric chelates 13 and 14 were prepared by adding a saturated solution of cupric chloride dihydrate, in absolute EtOH, dropwise to 10 (1.5 g) or 12 (1.0 g) in 20 mL of absolute EtOH until no additional precipitate formed. The reaction mixtures were stirred 15 min. The precipitates were collected, washed several times with triply distilled H20,and dried in vacuo.
(4) L. A. Cates, Y. M. Cho, L. K. Smith, L. Williams, and T. L. Lemke, J. Med. Chem., 19, 1133 (1976). ( 5 ) K. C. Agrawal, B. A. Booth, S. M. DeNuzzo, and A. C. Sartorelli, J . Med. Chem., 19, 1209 (1976), and references cited therein. (6) A. Leo, C. Hansch, and D. Elkins, Chem. Rev., 71,525 (1971). (7) 2[methylamine (-0.57) double bond (-0.30) + pyridine (+0.65)] diethyl phenylphosphorothioate (+3.46) - 2Et (+1.00 each) intramolecular hydrogen bonding (+0.65), given N and 0 are equivalent. (8) 7 - 2(pyridine) (+0.65) 2(pyridine N-oxide) (-1.69). (9) Methylamine (-0.57) + double bond (-0.30) + pyridine N-oxide (-1.69) + diethyl phenylphosphorate (+1.64) - 2Et ($1.00 each) intramolecular hydrogen bonding (+0.65), given N and 0 are equivalent. (10) L. A. Cates, M. B. Cramer, and L. Williams, J . Med. Chem., 21, 143 (1978). (11) Hexamethylphosphoric triamide (+0.28) - 2(branching of functional group) (-0.20 each) + Me (+0.50) + 2C1 (+0.39 each) + ring closure (-0.09) + intramolecular hydrogen bonding (+0.65). (12) Tripropylphosphine oxide (+2.53) - 2Me (+0.50 each) + 2C1 (+0.39 each) ring closure (-0.09) + intramolecular hydrogen bonding (t0.65). (13) B. A. Gingras, R. W. Hornal, and C. H. Bayley, Can. J. Chem., 38, 712 (1960). (14) G. J. Van Giessen and H. G. Petering, J . Med. Chem., 11, 695 (1968). (15) W. P. Purcell, G. E. Bass, and J. M. Clayton, “Strategy of Drug Design”, Wiley, New York, N.Y., 1973, p 127. (16) C. S. Hanes and F. A. Isherwood, Nature (London),164, 1107 (1949). (17) I. Berenblum and E. Chain, Biochem. J., 32, 295 (1938). (18) M. Nagazawa and Y. Imamiya, Japanese Patent 11013 (1962); Chem. Abstr., 59, 3833 (1963). (19) R. Klement and K. 0. Knollmuller, Chem. Ber., 93, 834 (1960). (20) H. Tolkmith, J. Am. Chem. SOC.,84, 2097 (1962). (21) E. P. Papadopoulos, A. Jarrar, and C. H. Issidorides, J . Org. Chem., 31, 615 (1966). (22) W. Autenrieth and 0. Hildebrand, Ber. Dtsch. Chem. Ges., 31, 1094 (1898). (23) F. Sondheimer, 0. Mancera, M. Urquiza, and G. Rosenkranz, J . Am. Chem. Soc., 77, 4145 (1955).
Acknowledgment. T h i s investigation was supported b y U.S. P u b l i c Health Service Grant CA-19882-01. The a u t h o r s wish t o t h a n k Norman Caron of the Department of Biophysical Sciences, University of Houston, for cond u c t i n g the Sarcoma 180 and P388 m u r i n e tests. References and Notes (1) A. C. Sartorelli, K. C. Agrawal, B. A. Booth, J. Pittman, D. Bartholomew, and A. D. Broom, J . Med. Chem., 19, 830 (1976). (2) K. C. Agrawal and A. C. Sartorelli, J . Med. Chem., 21, 218 (1978). (3) K. C. Agrawal, M. H. Lee, B. A. Booth, E. C. Moore, and A. C. Sartorelli, J . Med. Chem., 17, 934 (1974).
+
+
+
+
+
+
Synthesis of Spiro[isobenzofuran- 1(3H),4’-piperidines]as Potential Central Nervous System Agents. 4. Central Nervous System Depressants R i c h a r d C. Allen, Victor
J. Bauer, R a y m o n d W. Kosley, Jr., Arthur R. M c F a d d e n , Gregory M. Shutske,*
Chemical Research Department Michael L. Cornfeldt, S t u a r t Fielding, H a r r y M. Geyer, 111, and Jeffrey C. Wilker
Department of Pharmacology, Hoechst-Roussel Pharmaceuticals, Inc., Somerville, New Jersey 08876. Received May 26, 1978 The synthesis of 1’-[3-(4-fluorobenzoyl)propyl]-3-phenylspiro[isobenzofuran-l(3H),4’-piperidine] (2a) and eight halo and methoxy analogues is described. The compounds were generally more potent per os than chlorpromazine in the Sidman avoidance paradigm in rats and less potent than haloperidol. l’-[3-(4-Fluorobenzoyl)propyl]-3-(4fluorophenyl)spiro[isobenzofuran-l(3H),4’-piperidine] (2e) approached the per os potency of haloperidol in this test and was shown to be active in inhibiting monkey avoidance also. Compound 2e was much less active than haloperidol in antagonizing apomorphine-induced emesis in dogs, apomorphine-induced stereotypy in rats, and amphetamine-induced circling in lesioned rats. This lack of nonselective, dopamine-receptor blocking effects makes 2e attractive as a potential neuroleptic.
In t h e first t h r e e p a p e r s i n t h i s series’ we described a number of 3-phenylspiro[isobenzofuran-l(3H),4‘-piperidines] of general formula 1. C o m p o u n d s la and l b ex-
hibited potent antitetrabenazine activity and IC was active in lowering blood pressure i n the spontaneously hypertensive r a t . Since several c o m p o u n d s i n the 3-aryl-
0022-2623/78/1821-1149$01.00/0 0 1978 American Chemical Society
1150 Journal of Medicinal Chemistry, 1978, Val. 21, No. I 1 R
Allen et al. Scheme I q0ZCBH5
FH3 I
fN) I
A1
Ar
5a-i
4a-i
la, R = H b, R = OH c, R = SC,H,
Y
spiro[isobenzofuran-l(3H),4’-piperidine]series were candidates for development as antidepressants,larbefforts were made to capitalize on the ability of the unique parent heterocyclic ring system to form the basis for new central nervous system (CNS) drugs. Numerous alkyl, aralkyl, and heteroalkyl substituents were attached to the nitrogen of the spiro[isobenzofuran-l(3H),4’-piperidine] ring system with the aim of achieving compounds with CNS depressant activity. Of the compounds that were synthesized, the ones containing a 3-(4-fluorobenzoyl)propylmoiety, 2, showed R
I
~
1
x
X
& \
Ar
6a-i
71
,
/N\
I
1
I
I
Ar
2a-i xY f
l
H
3a, X = H; Y = C1
b , X = CF,;Y=H c, X = H; Y = Br 2, R2, R , = H, Cl, F, or OCH,
R, (CH2),COC,H,-p-F the most potential as CNS depressants. It was recognized that these compounds are structurally similar to some known neuroleptics-haloperidol (3a), trifluperidol (3b), and bromoperidol (3c). This paper describes the synthesis and pharmacology of nine such 3-arylspiro[isobenzofuran-l(3H),4’-piperidine] derivatives, 2. Chemistry. The synthesis of compounds 2 is described in Scheme I and their properties are given in Table I. The properties of the novel intermediates are given in Table 11. The N-methyl compounds 4a-d,hla and 4e-g2 are described in previous publications. Compound 4i was synthesized by the addition of phenyllithium to 6chloro- 1’-methylspiro[ isobenzofuran-1(3H),4’-piperidin] %one (7)3 and subsequent reduction of the 6-chloro-1’methyl-3-phenylspiro[isobenzofuran-l(3H),4’-piperidin]3-01 (8) with formic acid.? Treatment of 4a-i with phenyl chloroformate in dichloromethane gave phenyl carbamates 5a-i which were hydrolyzed with potassium hydroxide in ethylene glycol to give the secondary amines 6a-i. Alkylation of the secondary amines with 2-(3-chloropropyl)-2-(4-fluorophenyl)-1,3-dioxolane in refluxing 1butanol containing potassium carbonate, followed by hydrolysis of the ketals, gave the 3-arylspiro[isobenzofuran-1(3H),4’-piperidines] 2a-i. Pharmacology. The selective inhibition of a nondiscriminated avoidance response is a property common to neuroleptics, anxiolytics, and sedatives. The compounds in this series were tested in rats in the operant paradigm originally described by Sidman4and elegantly used by the Janssen group’ to develop compounds, including 3a-c. In this test, neuroleptics such as chlorpromazine or halo-
8
4i
peridol will suppress a learned response that an animal has been trained to employ in order to avoid an unsignaled, unpleasant stimulus. Spiro[isobenzofuran-l(3H),4’piperidines] 2a-i (Table 111) were generally more potent PO than chlorpromazine and less potent than haloperidol in this test. They exhibited a delayed onset, with peak activity occurring 4-6 h after dosing. They also were often more potent PO than ip. In these cases PO EDjovalues were calculated a t the time of maximum effect. In general, the substituted analogues were more potent than the unsubstituted parent, 2a. Within the methoxy analogues, 4’-substitution gave the most active compound, 2d, with an EDjo of 3.0 mg/kg PO, while the 5-substituted analogue 2b and the 6-substituted analogue 2c were less than half as potent. The fluoro analogues 2e-g were approximately equipotent, with the 4’,6-difluoro compound 2g being the most active compound in inhibiting Sidman avoidance in rats (ED,, = 2.2 mg/kg PO). The chloro analogues presented some striking differences. Although the 6-chloro and 6-fluor0 analogues 2i and 2f were approximately equipotent, the 4’-chloro analogue 2h, in sharp contrast to the 4’-flUOr0 analogue 2e, was only weakly active. Neuroleptics can be distinguished from other CNS depressants in a conditioned avoidance-escape procedure such as the pole-climb avoidance paradigmsfiIn this test an animal is conditioned to respond (pole climb) to a stimulus (light and tone) in order to avoid an unpleasant foot shock. Should it fail to give the proper response in
Journal of Medicinal Chemistry, 1978, Vol. 21, No. 11 1151
Spiro[isobenzofuran-l(3H),4'-piperidinesl
-3-arylspiro[ isobenzofuran-l(3H),4'-piperidinesIo Table I. 1'-[3-(4-Fluorobenzoyl)propyl]
*e s\
Ar
comDd
X
2a 2b 2C 2d 2e 2f 2g 2h 2i
H 5-CH30 6-CH30 H H
Ar
mp,b,e C
yield,c %
4-CH30C,H, 4-FC6H,
137-138 62-64 98-100 11 3-11 4 132-1 34 116-118 113-115 131-133 90-92
88 53 55 66 58 88 65 64 43
C6H5 C6H5
C6H5
6-F 6-F
C6H5
H
4-FC6H, 4-C1C6H,
6431
C6H5
formulad CZ*HZ,FNO2 C29H30FN03
C29H30FN03 cZ9
H3 OFN03
CZ 8H,,F,NO, C28H27F2NOz 2'
8 H 2 6 F3N02
C, 8H2,ClFNO, C2,H2,C1FNOz
Melting points are uncorAll compounds exhibited IR and ' H NMR spectra consistent with the assigned structures. All compounds analyzed correctly for C, H, Yield of analytically pure material; yields were not optimized. rected. and N within * 0.4% of theoretical values. e All compounds were recrystallized from acetonitrile. Table 11. Novel Spiro[ isobenzofuran-l(3H),4'-piperidine] Intermediates'
8
X s\
Ar
compd
4i 5c 5e
6f
5n 6i 6f
I
6i 8
Rl CH, C02C6H5
CO,C,H, CO;C;H; CO, C.H. CO;C,H; H H CH,
X 6-C1 6-CH30 H 6-F 6-F 641
6-F 641 641
See corremondine footnotes t o Table I. acetate; F = hexane. -
a-d
R2
mp,b O c
H H H H H H H H OH
117-118 161-162 181-183 147-149 137-139 159-160 129-131 117-117.5 208-210
Ar 6'
HS
C6H5
4-FC,H. C6Hs" 4-FC,HA C,H,"
-
6' HS C6H5 C6HS
e
R2
yield,c recrystn % solvente
85 72 79 70 76 84 80 92 75
F B B C-E A-B C
F F
B-D
formulad C,,H, ,C1NO C,6H2
SN04
C, FNO, C;;H;;FNO; C,.H,,F,NO, C;;H;;CiNO; C,*Hl,FNO C,,H,,ClNO C,,H,,ClNO,
A = water; B = acetonitrile; C = methanol; D = tetrahydrofuran; E = ethyl
order to avoid the shock, however, the pole is still available so that it can escape the shock once it is delivered. Typical neuroleptics show a wide separation between the dose that will suppress conditioned avoidance responding and the dose that will suppress the escape response. Sedatives and anxiolytics, on the other hand, show much less separation (see Table 111). The compounds of this series which were active in inhibiting Sidman avoidance all showed large differences between the EDso values for suppression of avoidance responding and for escape failures (Table 111). Compounds 2a,b,e were roughly as potent as chlorpromazine in suppressing avoidance in this test but caused escape failures a t higher doses. Compound 2f approached the potency of haloperidol in suppressing conditioned avoidance, again requiring a higher dose for escape failure, The weak activity of the 4'-chloro analogue 2h was confirmed in this test. Compounds 2a,b,e,f showed consistent activity in inhibiting Sidman and pole-climb avoidance in rats. Table IV shows that the activity of 2b and 2e was maintained in suppressing Sidman avoidance in monkeys. The antiavoidance effect of 2e was partially reversed by
the anticholinergic agent, benztropine. This provided another indication of true neuroleptic activity.' It is thought that antipsychotic agents exert their effects by dopamine-receptor blockade in the CNS? particularly in the ventral limbic areas,B Furthermore, it is postulated that the troublesome Parkinsonian-like extrapyramidal (EPS) side effects of the clinically useful antipsychotics are due to nonselective, dopamine-receptor blockade in the corpus s t r i a t ~ m .Pharmacologically, ~ it is observed that typical antipsychotics like chlorpromazine and haloperidol block the effects of dopamine agonists in a nonspecific manner. For example, they inhibit apomorphine-induced emesis in dogs and apomorphine-stereotyped behavior in rats.l0 In view of this, it is interesting to note (Table 111) that compounds 2b,e,g were a t least ten times less potent than haloperidol in blocking apomorphine emesis in dogs. It should be particularly noted that compound 2e approached the PO potency of haloperidol in Sidman avoidance and yet was a t least two orders of magnitude less potent in blocking apomorphine emesis. This lack of antiapomorphine effects was seen in rat stereotypy also.
1152 Journal of Medicinal Chemistry, 1978, Vol. 21, No. 11
Allen et al.
Table 111. CNS Activity of 1'-[ 3-(4-Fluorobenzoyl)propyl]-3-arylspiro[ isobenzofuran-l(3H),4'-piperidines]a,b pole-climb avoidance, ED,,, mg/kg ip Sidman avoidance, (rat) ED,,, m d k g compd ip (rat) A R ~ EF' 2a 6.2 (5.5-7.0) 8.6 (6.8-10.1) >40 4.8 (4.5-5.1)" 2b 9.5 (8.8-10.4) 9.8 (7.9-11.3) >40 2c >loc ntd e 2d 3.0 (2.7-3.3)" 17.2 (15.7-18.8) >40 2e 3.7 (3.4-4.3)' 4.3 (3.7-4.8)' 16.2 (13.6-20.6)' 2f 2.7 (2.3-3.2) 0.95 (0.66-1.25) 31.1 (20.1-63.9) 2g 2.2 (1.9-2.5)' 18.6 (15.8-22.4) >40 2h >10 > 20 e h e 2i 2.3 (1.9-3.0) haloperidol 0.35 (0.30-0.40)' 0.53 (0.44-0.64) 7.3 (6.1-9.7) chlorpromazine 2.3 (2.0-2.7) 5.2 (4.7-5.7) 9.0 (8.2-10.1) 7.5 (7.2-7.8)' 13.8 (10.9-16.5) >60 thioridazine .i pentobarbital 23.7 (21.5-27.0) 24.1 (23.0-25.4) diazepam 13.4 (11.0-15.8) 24.5 (21.9-30.0)
apomorphine emesis, ED,,, mg/kg PO (dog) 0.15 (0.13-0.18) >1.0 nt 0.16 (0.13-0.22) >5.0 0.047 (0.033-0.098) 0.5f >10 0.05 (0.04-0.14) 0.06 (0.03-0.14) 12.8 (7.8-16.6)
amphetamineinduced circling, ED,,, mg/kg ip (rat) 2.6 (1.9-3.3) 9.9 (8.9-10.9) nt 10.2 (9.2-11.5) 11.9 (10.2-14.4 4.6 (3.9-5.5) 9.7 (8.1-11.9) 6
3.5 (3.0-4.3) 0.10 (0.07-0.12 2.2 (2.0-2.5) 13.3 (11.3-16.1)
a Figures in parentheses are 95% confidence limits. Test procedures are described in the Experimental Section. Administered PO. Not tested. e Not calculated. A dose of 0.5 mg/kg produced a 50% decrease in emesis, but no ED,, was calculated. g A 47% increase in circling occurred at 5 mg/kg. A dose of 5 mg/kg produced a 47% decrease in A PO dose of 1.25 mg/k produced a 50% reduction in responding, but n o ED,, was calculated. avoidance responding. I A dose of 20 mg/kg produced 1 5 % inhibition. AR = avoidance response. E F = escape failure.
a
'
Table IV. Inhibition of Sidman Avoidance in Monkeysa decrease in responses, % compd 2b 2c 2e chlorpromazine haloperidol
lb
2
3
39 84 90 75 32 54 19 91 9
42 77 92 17 21 83 29 97 68
29
dose, 4 mg/kgpo 76 2.5
-
-
90
-
-
70 48 30 2
Table v. Inhibition of Apomorphine and Amphetamine Stereo typic@
68 11 97 26
10 20' 10 5
5.0 2.5 1.0 0.75
a The test procedure is described in the Experimental Section. A maximum of four monkeys was tested at a given dose. A "-" indicates that an animal was not tested at a given dose. ' This effect was partially reversed by benztropine (1.25 mg/kg PO).
Stereotyped behavior (compulsive licking, gnawing, and sniffing) in rats depends on a direct stimulation of receptors for dopamine in the CNS." Antagonism of this behavior is a useful, if nonspecific, in vivo test for dopamine-receptor blockkg activity. Apomorphine initiates stereotypy by directly stimulating dopamine receptors. Amphetamine, however, acts in an indirect manner by stimulating dopamine release, leading to stereotyped behavior." Chlorpromazine and haloperidol antagonize both apomorphine- and amphetamine-induced stereotypy. Compounds 2b,e, however, besides being much less active than haloperidol in antagonizing apomorphine-induced emesis in dogs, were inactive in blocking apomorphineinduced stereotypy in rats at the doses tested (Table V). They were, however, moderately active in antagonizing amphetamine-induced behavior. Thus, the operant avoidance blocking behavior of compounds 2b,e indicated their potential as neuroleptics, while their inability to antagonize apomorphine in vivo showed a lack of nonspecific, dopamine-receptor blocking effects. These data suggested that these compounds could be of potential antipsychotic utility in man but without the undesirable EPS reactions resulting from nonselective, dopamine-receptor blockade. This view was further strengthened by the ED,, values of 9.9 and 11.9 mg/kg ip that 2b and 2e displayed, re-
compd 2b 2e haloperidol chlorpromazine thioridaz ine
inhibn of apomorphine stereotypy, ED,,, m g / k sc (rat ) C C
0.20 (0.19-0.21) 5.4 (1.8-5.7) ntd
inhibn of amphetamine stereotypy, ED,,, mg/kg ip (rat) 6.97 (5.11-9.52) 11.8 (3.4-14.0) 0 . 2 1 (0.11-0.34) 1.7 (1.1-3.2) 15.8 (6.1-32.7)
a Figures in parentheses are 95% confidence limits. The test procedure is described in the Experimental Section. ' Inactive at 25 mg/kg. Not tested.
spectively, in inhibiting amphetamine-induced circling in lesioned rats (Table 111). This test gives an indication of direct, dopamine-receptor blockade in the basal ganglia and, as such, may predict the EPS side effect liability of potential antipsychotics.'* Haloperidol, for example, is particularly effective (EDN= 0.10 mg/kg ip) in inhibiting this circling behavior. Thioridazine, however, a compound with less E P S liability,13 has an ED,, of 13.3 mg/kg ip. Compounds 2b,e thus fall in the range of thioridazine. By balancing conditioned avoidance behavior against undesirable dopamine-receptor blocking effects, the 4'fluoro analogue 2e was chosen for further development as a potential neuroleptic with a unique pharmacological profile. Experimental Section The structures of all compounds are supported by their IR (Perkin-Elmer 457) and 'H NMR (Jeolco CGOHL, tetramethylsilane) spectra. Melting points were taken on a Thomas-Hoover capillary melting point apparatus. All melting points are uncorrected. Elemental analyses were performed by Micro-Tech Labs, Skokie, Ill. The synthesis of starting materials 4a-d,h1* and 4e-g2 is described elsewhere. The properties of the novel compounds are described in Tables I and 11. 6-Chloro-1'-methylspiro[ isobenzofuran- 1(3 H),4'-piperidin1-3-one (7). Compound 7 was prepared as described by Rodriguez et al. in 40% yield: mp 214-216 "C (lit.3mp 209-213 "C). 6-Chloro-l'-methyl-3-phenylspiro[isobenzofuran-l(3H),4'-piperidin]-3-01 (8). To a stirred solution of 260 mL of
Spiro[isobenzofuran-l(3H),4’-piperidines]
Journal of Medicinal Chemistry, 1978, Vol. 21, No. 11 1153
2.3 N phenyllithium (0.6mol) in 70:30 benzene/ether was added a warm solution of 7 (50.3g, 0.2mol) in 1 L of T H F and 400 mL of toluene. An ice bath was applied occasionally to keep the reaction temperature below 40 “C. After the addition was complete, the reaction was stirred for 3 h and then quenched with 200 mL of HzO. The organic phase was separated, washed with additional HzO, dried, and concentrated in vacuo to give crystalline 8. This was filtered and washed with EtzO to give 49.5g (75%) of 8. 6-Chloro-l’-methyl-3-phenylspiro[ isobenzofuran-l(3H),4’-piperidine] (4i). As described for compounds 4e-g,’ 8 (47.5g, 0.144 mol) was treated with 400 mL of refluxing 98% formic acid to give 38.6 g (85%) of 4i. 6-Chloro-1’-(phenoxycarbonyl)-3-phenylspiro[isobenzofuran-l(3H),4’-piperidine] (59. To a solution of 4i (34.8g, 0.11 mol) in 200 mL of CH2C12was added 20 g (0.13mol) of phenyl chloroformate in 200 mL of CHzClz. After 16 h the solvent was removed in vacuo and the resulting solid triturated with EtzO to give 38.8 g (84%) of 5i. Compounds 5a-h were prepared analogously. 6-Chloro-3-phenylspiro[isobenzofuran-1(3H),4’-piperidine] (6i). Ethylene glycol (800 mL) was added to a freshly prepared solution of 74 g of potassium hydroxide pellets in 35 mL of HzO. Compound 5i (32.1 g, 0.076 mol) was then added and the resulting suspension brought to reflux with vigorous stirring. After refluxing for 30 min the solution (now homogeneous) was poured into 2.5 L of ice and H20, and the precipitated 6i was filtered off. This material was precipitated once from MeOH and HzO to give 21 g (91%) of 6i. Compounds 6a-h were prepared analogously. 6-Chloro-l’-[3-(4-fluorobenzoyl)propyl]-3-phenylspiro[isobenzofuran-l(3H),4’-piperidine] (29. Compound 6i (8.0 g, 0.027 mol) was dissolved in 50 mL of 1-butanol and 15 g of anhydrous KzCO3 was added. This suspension was brought to reflux and 2-(3-chloropropyl)-2-(4-fluorophenyl)-1,3-dioxolane (9.0 g, 0.037 mol) was added dropwise. After refluxing for 20 h, the inorganic salts were filtered off and the solvent was removed in vacuo. The crude ketal thus obtained was then hydrolyzed by boiling briefly in 300 mL of MeOH and 150 mL of 3 N HC1. This solution was chilled, washed with hexane, made basic, and extracted with CHC13. The dried organic extracts were concentrated in vacuo to give an oil that crystallized from CH3CN to give 5.34 g (43%) of 2i. Compounds 2a-h were prepared in a similar manner. Pharmacological Methods. Compounds were administered using distilled water and a suitable surfactant except where otherwise indicated. EDSo values were calculated during a representative time of peak activity using computerized linear regression analysis. Sidman Avoidance in Rats. In an apparatus similar to that described by Janssen? male Sprague-Dawley rats (Charles River) were trained to avoid an unsignaled shock by repetitive leverpressing responses. A shock-shock interval of 15 s and a response-shock interval of 30 s were used. A drug’s effect on the performance of each animal was compared to the performance data generated in the previous nondrug session. Each animal thereby served as its own control. Responses constituted the basic measure of performance during a specific time interval and were reported as percent of control responses. The ED50 of a test compound was that dose which caused a 50% reduction in responding at the time of maximum effect. Sidman Avoidance in Monkeys. Male squirrel monkeys (Sairnirisciureus) were trained as described above for rats. A shock-shock interval of 20 s and a response-shock interval of 20 s were used. The basic data were accumulated as with rats over a 4-h test session. The results were reported as a percent decrease in responses. Pole-Climb Avoidance in Rats. Male Long-Evans rats (Blue Spruce Farms) were trained and tested in experimental chambers similar to those described by Cook,6 with the following modifications. On the top of the experimental chamber a speaker and a light were situated. A smooth stainless steel pole, 1 in. in diameter, was suspended by a spring through a hole in the center of the chamber top. The pole could be pulled down approximately in. by a weight greater than 200 g to activate a microswitch. When a rat jumped on the pole and activated the microswitch,
a response was recorded. Due to the spring tension, a rat could not hold the pole down while standing on the grid floor and because of the smooth surface it could not remain on the pole any length of time. The activation of both light and speaker together was used as the conditioning stimulus (CS). The CS was presented alone for 5 s and for the following 30 s the CS was coincident with a scrambled shock from the grid floor, the unconditioned stimulus (UCS). A response (pole jump) during the CS period terminated the CS and the following UCS CS. This was considered an avoidance response (AR). Response when both CS and UCS were present terminated both stimuli and was an escape response (ER). Although all stimuli were terminated by AR and ER, their respective clocks maintained the prescribed durations and in effect the intertrial interval (ITI) was expanded. A response during this IT1 time was not recorded and had no effect on the system. The rats which consistently reached or exceeded 80% avoidance without any escape failures were used with experimental compounds. The animals were then compared with their solvent control data in avoidance responses and escape failures (EF). Two EDMvalues were calculated, representing, respectively, the dose that caused a 50% decrease in AR and the dose that caused a 50% increase in E F at the time of maximum effect. Apomorphine-Induced Emesis in Dogs. Adult beagle dogs of either sex were given the test compound PO as a mixture with lactose in a gelatin capsule; they were then dosed with a standard 0.15mg/kg dose of apomorphine hydrochloride subcutaneously at various intervals after administration of the test compound. The initial screen was usually a dose of 1 mg/kg PO of the test compound. The dogs were first observed for overt behavioral effects, e.g., pupillary response to light, changes in salivation, sedation, tremors, etc. After the administration of apomorphine the dogs were observed for stereotypic sniffing and gnawing and for the emetic response. Emesis was defined as wretching movements followed by an opening of the mouth and either attempted or successful ejection of stomach content. The ED50 was defined as the dose that produced 50% inhibition in this emetic response a t the time of maximum effect. Stereotypy in Rats. Groups of male Wistar rats were given the drug 1h prior to apomorphine or amphetamine challenge and were placed in individual stainless steel cages. A control group received the vehicle. The rata were challenged with apomorphine hydrochloride (Merck; 1.25 mg/kg iv) or d-amphetamine sulfate (Smith Kline & French Labs; 10 mg/kg iv) and then observed a t 5, 10, and 20 min after apomorphine challenge or 55 and 65 min after amphetamine challenge. Stereotypic activity was defined as sniffing, licking, or chewing behavior that occurred in a repetitive manner and was noted as follows: 0 = no unusual activity; 1 = occasional sniffing, licking, or chewing but with periods of normal behavior longer than periods of stereotypy; 2 = frequent sniffing, licking, or chewing but with occasional brief periods of normal behavior; 3 = constant sniffing, licking, or chewing without interruption, compulsive gnawing possibly present. The scores for stereotypy were added and averaged for each animal. The percent inhibition shown by the drug was calculated and an EDSowas determined as the dose required to produce 50% inhibition of stereotypy. Rat Circling. Male Wistar rats were anesthetized (sodium pentobarbital, 18 mg/kg ip) and the head was placed in a stereotaxic device and positioned in the plane of the Koenig and Klippel atlas. A radio frequency electrode was aimed at the zona compacta of the substantia nigra (coordinated anterior 2.18,lateral 2.00, and dorsal ventral 2.0 from instrument zero) through a hole drilled in the exposed skull. The lesion was made using a radio frequency lesion generator at a temperature of 50-55 OC for 45 s. The electrode was then withdrawn; the animal was sutured and allowed a 1-week recovery period. Specially constructed opaque plastic spheres attached to solid-state programming equipment served as test chambers. The number of full turns ipsilateral to the lesion was recorded on an automatic printout counter every 15 min for a 2-h test period. To determine the control values for turning, each subject was injected with 5 mg/kg ip of d-amphetamine and immediately placed in the chamber on two successive periods separated by a week.
+
Temussi, Lelj, Tancredi
1154 Journal of Medicinal Chemistry, 1978, Vol. 21, No. 11
Test compounds were typically administered 60 min prior to challenge by 5 mg/kg ip of amphetamine. The ED,, was calculated as the dose that produced a 50% decrease from control values in amphetamine-induced circlings. Acknowledgment. The authors wish to express their appreciation to Marc N. Agnew and David L. Clare for spectral data and to Susan A. Bullock, Craig Dobson, James G. Kiley, Jr., Kathleen M. Morrison, Daniel Salomone, Mark R. Szewczak, and Richard D. Warren for performing pharmacological assays. We also gratefully acknowledge June B. Baird and Ann Van Dine for library research and Irene H. Field for assistance in preparation of the manuscript. References and Notes (1) (a) For paper 1 of this series, see V. J. Bauer, B. J. Duffy, D. Hoffman, S.S.Klioze, R. W. Kosley, Jr., A. R. McFadden, L. L. Martin, H. H. Ong, and H. M. Geyer, III., J . Med. Chem., 19, 1315 (1976); (b) for paper 2, see S. S.Klioze, V. J. Bauer, and H. M. Geyer, 111, ibid., 20,610 (1977); (c) for paper 3, see S.S.Klioze and W. J. Novick, Jr., ibid., 21,
400 (1978). (2) G. M. Shutske, J. Org. Chem., 42, 374 (1977). (3) A. Marxer, H. R. Rodriguez, J. M. McKenna, and H. M. Tsai, J . Org. Chem., 40, 1427 (1975). (4) (a) M. Sidman, Science, 118, 157 (1953); (b) Ann. N.Y. Acad. Sci., 65, 282 (1956). ( 5 ) C. J. E. Niemegeers, F. J. Verbruggen, and P. A. J. Janssen, Psychopharmacologia, 16, 161 (1969). (6) (a) L. Cook and E. Weidley, Ann. N.Y. Acad. Sci., 66, 740 (1957); (b) L. Cook and A. C. Catania, Fed. Proc., Fed. Am. SOC.E x p . Biol., 23, 818 (1964). (7) H. M. Hanson, C. A. Stone, and J. J. Witoslawski, J . Pharrnacol. E x p . Ther., 173, 117 (1970). (8) A. Carlsson, J . Psychiatr. Res., 11, 57 (1974). (9) N.-E. Anden, J . Psychiatr. Res., 11, 97 (1974). (10) S. Fielding and H. Lal, “Industrial Pharmacology, Neuroleptics”,Vol. 1, S.Fielding and H. Lal, Ed., Futura Publishing Co., Mount Kisco, N.Y., 1974, p 64. (11) A. M. Emst, Psychopharmacologia, 10, 316 (1967). (12) R. J. Stawarz, H. Hill, S. E. Robinson, P. Setler, J. V. Dingell, and F. Sulser, Psychopharmacologia, 43, 123 (1975). (13) J. 0. Cole and D. J. Clyde, Rev. Can. Biol., 20, 565 (1961).
Three-Dimensional Mapping of the Sweet Taste Receptor Site Piero A. Temussi,* Istituto Chimico, Uniuersity of Naples, 80134 Naples, Italy
Francesco Lelj, Dipartimento d i Chimica, University o f Calabria, Cosenza, Italy
and Teodorico Tancredi L.C.M.I.B. of C.N.R., Arco Felice, Naples, Italy. Received February 6, 1978
The active sites of the receptors for sweet and bitter tastants are shown to be related by a simple symmetry operation. This relationship, in turn, allows the identification of the critical geometrical features of both receptor sites. The model proposed for the sweet site is shown to be consistent with a large number of (conformationallyrigid) sweet mo1ecuies.-
Two of the so-called primary taste qualities,’ i.e., sweet and bitter, can be elicited by a wide variety of apparently unrelated molecules.’%2It is the purpose of the present work to show that the active sites of sweet and bitter receptors are related by a simple symmetry operation and that it is possible to find the main geometrical features of both sites by using the shapes of conformationally rigid molecules as idealized molds. Although the existence of common features among tastants had been suspected long ago, in relation with the theory of selective adsorption3 of membrane proteins, it has been only in comparatively recent times that one of these features has been actually identified. Shallenberger and Acree4 have shown that a characteristic common to nearly all known sweet substances is an entity composed of a basic (B) and acidic (AH) group spaced of about 0.3 nm that should interact with a similar (complementary) entity of the receptor site, possibly through a pair of hydrogen bonds. Shallenberger has also shown5 that the drastic differences in taste existing between amino acid enantiomers may be explained by the presence in the receptor site of a spatial barrier located a t approximately 0.3 nm from the line joining AH and B. That is, sweet amino acids must have an R configuration since their side chain would otherwise “invade” the spatial barrier. These facts (which henceforth shall be referred to as “Shallenberger’s theory“) do not account satisfactorily for all aspects of sweet taste. In particular, Shallenberger’s 0022-2623/78/1821-1154$01.00/0
theory gives no clues for the understanding of the large differences in relative sweetness among known tastants and does not explain why substances with the right AH-B entity (such as simple geometric isomers of known sweeteners) are completely tasteless. A dramatic example of this type is afforded by the two isomers of the oxime of anisaldehyde; the anti isomer is very sweet but the syn isomer is tasteless.2 A trivial way to overcome these difficulties may be to invoke the existence of multiple (specialized) sites for a given taste sensation. Evidence in favor of the existence of multiple sites both for the bitter and sweet taste has, in fact, been given by various author^.^^^ On the other hand, rather than attributing every peculiarity to an increasing number of distinct receptors it is certainly more fruitful to deepen the analysis of the properties of a given receptor site, trying to explain the taste of the greatest possible number of molecules in terms of a single site if it can be clearly identified. In fact, the number of sweet molecules described by Shallenberger4in terms of the AH-B entity is large enough to justify the reference to a common receptor site. For the sake of clarity it may be convenient to classify the features of the active site as electronic or geometric although in some instances such a classification can appear somewhat artificial. The main “electronic” feature of this site is obviously represented by the AH-B entity itself, whereas the spatial barrier represents a simple geometric feature. It is the purpose of this paper to show how this view of 6 1978 American Chemical Society
